Previously, in studies involving a semi-homologous system of gnotobiotic newborn pigs and a virulent porcine rotavirus strain (SB- 1A) and an avirulent human rotavirus strain (DS-1) and their reassortants, we demonstrated that: (i) the third (VP3), fourth (VP4), ninth (VP7), or tenth (NSP4) porcine rotavirus gene each play an important independent role in the virulence of rotavirus infection in piglets; and (ii) all four of the porcine rotavirus virulence- associated genes are required for the induction of diarrhea and the shedding of rotavirus by piglets. These observations suggested a potential new strategy for attenuation of wild-type human rotaviruses of major epidemiological importance and its application to the development of a safe and effective vaccine. Using this strategy, we were successful in generating four double gene substitution human x bovine rotavirus reassortants, each of which possessed the following: (i) the VP4-encoding gene from human rotavirus Wa (VP4:1A); (ii) the VP7-encoding gene from human rotavirus D (VP7:1), DS-1 (VP7:2), P (VP7:3) or ST3 (VP7:4); and (iii) the remaining nine genes including the VP3-encoding gene and NSP4-encoding gene from bovine rotavirus UK. In addition, we successfully generated (i)three human x bovine rotavirus single gene substitution reassortants, each of which had the VP4-encoding gene from human rotavirus Wa (VP4:1A), Mo (VP4:1A) or DS-1 (VP4:1B) and the remaining genes from bovine rotavirus UK and (ii) two human x rhesus rotavirus single gene substitution reassortants, each of which had the VP4-encoding gene from human rotavirus Wa or DS-1 and the remaining genes from rhesus rotavirus MMU18006. Recently, unusual VP7 (G) serotypes causing a high incidence of human infection have been detected in various parts of the world including G5 strains in Brazil, G8 strains in Central Africa, G9 strains in India and Bangladesh, and G10 strains in India. Because of the possibility in the future that such G serotypes might need to be included in our vaccine candidates and therefore to ?be prepared?, we generated three rhesus x human and three bovine x human rotavirus reassortants, each of which had only one gene from human rotavirus encoding G5 or G9 specificity or bovine rotavirus encoding G10 specificity, and the remaining genes from either rhesus rotavirus MMU18006 or bovine rotavirus UK. Such strains may prove to be important for the development of an optimally effective rotavirus vaccine. - rotavirus, infantile diarrhea, vaccine, pathogenesis